From: Ioannis Giotis (giotis@cs.washington.edu)
Date: Mon Oct 04 2004 - 07:29:58 PDT
THE DESIGN PHILOSOPHY OF THE DARPA INTERNET PROTOCOLS
David D. Clark
The author talks about the history and introduction of the internet. He
focuses mainly on the intuition and the important design decisions that had
to be taken during the early stages of development. The main goals that led
to TCP/IP's current form are first presented and the author does a good job
in explaining and justifying the design while at the same time acknowledges
the limitations and problems that arise from the specific design.
It is surprising to see that even in a 1988 paper some of the most discussed
issues associated with the current form of the TCP/IP layers or the
underlying datagram were well recognized. The use of interesting examples
makes it more apparent that the author had a good idea as to where things
are going from there and future issues. It should also be noted that the
author does a good job on equally presenting issues from TCP,IP,datagram
that give a complete picture about the design but also allow us to see how
one layer's design influenced (or limited) the underlying layer.
On the other hand, while some justification exists, it is unclear to me why
the design choices made were the optimal ones; I would like to see other
alternative ideas that also met the goal criteria but were rejected for some
reason or even better some real-world benchmarks comparing performance of
different solutions. Of course, the main problem of this paper in the
current context, is its relevance to today's use of the internet. Despite
the specific example of VoIP, the significance of those applications is much
greater now and the arguments used probably need re-evaluation. A more
recent paper would also be more involved in the application layer design.
Overall, it was a nice read and achieved its goal of presenting the
motivation behind the implementation.
This archive was generated by hypermail 2.1.6 : Mon Oct 04 2004 - 07:29:58 PDT